diagnosis of peritonitis using near-infrared optical ... filediagnosis of peritonitis using...

Diagnosis of peritonitis using near-infrared opticalimaging of in vivo labeled monocytes-macrophages

Marcus-René LisyElisabeth SchülerFriedrich Schiller University JenaInstitute of Diagnostic and Interventional

RadiologyFZL Erlanger Allee 101D-07747 Jena, Germany

Frank LehmannPeter CzerneyDyomics GmbHWinzerlaer Str. 2AD-07745 Jena, Germany

Werner A. KaiserIngrid HilgerFriedrich Schiller University JenaInstitute of Diagnostic and Interventional

RadiologyFZL Erlanger Allee 101D-07747 Jena, Germany

Abstract. Peritonitis is an inflammatory process characterized bymassive monocytes-macrophages infiltration. Since early diagnosis isimportant for a successful therapeutic outcome, the feasibility for aselective labeling and imaging of macrophages for highly sensitiveoptical imaging was assessed. After in vitro incubation of mouse mac-rophages J774A.1 with the far-red/near-infrared fluorochrome DY-676, distinct fluorescence intensities �1026±142 a.u.� were detectedas compared to controls �552±54 a.u.� using a whole-body smallanimal near-infrared fluorescence �NIRF� imaging system. Macro-phage labeling was confirmed by confocal laser scanning microscopy�CLSM� and fluorescence-activated cell sorting, �FACS�. The fluoro-chrome was also found to be predominantly distributed within com-partments in the cytoplasm. Additionally, peritonitis was induced inmice by intraperitoneal injection of zymosanA. After intravenous in-jection of fluorochrome �55 nmol/kg� and using whole-body fluores-cence imaging, higher fluorescence intensities �869±151 a.u.� weredetected in the peritoneal area of diseased mice as compared to con-trols �188±41 a.u.�. Furthermore, cells isolated from peritoneallavage revealed the presence of labeled monocytes-macrophages. Theresults indicate that in vivo diagnosis of peritonitis by near-infraredoptical imaging of labeled monocytes-macrophages is feasible. Possi-bly, early stages of other inflammatory diseases could also be detectedby the proposed diagnostic method in the long term. © 2006 Society ofPhoto-Optical Instrumentation Engineers. �DOI: 10.1117/1.2409310�

Keywords: inflammation; peritonitis; diagnosis; contrast agent; near-infraredfluorescence; optical imaging; inflammation.Paper 06005R received Jan. 13, 2006; revised manuscript received Aug. 4, 2006;accepted for publication Aug. 10, 2006; published online Jan. 2, 2007.

1 Introduction

Peritonitis is an inflammatory process that can arise from vari-ous causes, including infectious, chemical, iatrogenic, andother etiologies. Patients with peritonitis have fever and sufferfrom abdominal pain and emesis. Various reasons are oftendescribed to be responsible for the disease, for instance, theinfection of abdominal organs �e.g., the pancreas� or injuriesduring surgical interventions.1,2 Even starch or talc from sur-gical gloves may provoke an abdominal inflammation.3 Inwomen, peritonitis often occurs after bacterial infection of theuterus.4 Moreover, patients with kidney failures treated byperitoneal dialysis commonly suffer from this disease.5

If untreated, there are many complications resulting fromperitonitis that may emerge quite rapidly, such as kidney,lung, and liver failures.6 Nevertheless, the current diagnosticmethods are based on the symptoms and are limited to theapplication of invasive procedures such as abdominal para-centesis for the assessment of the abdominal fluid in relation

to the presence of germs.7,8 Using radiologic methods, x-raydetections can suggest the presence of the condition, but thediagnostic potential is limited to anatomical information.9 Oc-casionally, magnetic resonance tomography �MRT� is used forabdominal imaging and can provide information about in-flammatory processes in this cavity. Nevertheless, early detec-tion is not possible by using MRT, particularly because itlacks sensitivity in comparison to other progressive tech-niques, like positron emission tomography �PET� in nuclearmedicine, which is based on radioactive probes as cellular ormolecular markers.

At the cellular level, peritonitis, like other inflammatoryprocesses, is characterized by a massive leukocyteinfiltration,10 such as activated monocytes-macrophages intothe target tissue 6 to 24 h after onset of inflammation.11,12 Ac-tivation of the monocytes-macrophages population takes placeat the highly vascularized internal mucosa layers,13 wherethose cells could theoretically be effectively labeled with theaforementioned probes for imaging purposes. The use of far-red and near-infrared fluorescence imaging for the highly sen-sitive detection of changes on the molecular and cellular levelis very attractive. Benefited by the increased capacity of light

1083-3668/2006/11�6�/064014/9/$22.00 © 2006 SPIE

Address all correspondence to Dr. Marcus-René Lisy, Friedrich Schiller Univer-sity of Jena, Institute of Diagnostic and Interventional Radiology, Dept. Experi-mental Radiology, FZL Erlanger Allee 101, 07747 Jena, Germany. Tel: 0049-3641-9325924; Fax: 0049-3641-9325922; E-mail: [email protected].

Journal of Biomedical Optics 11�6�, 064014 �November/December 2006�

Journal of Biomedical Optics November/December 2006 � Vol. 11�6�064014-1

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 31 Mar 2019Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

at wavelengths between 600 and 950 nm to penetrate tissuesup to a few centimeters,14,15 fluorochromes, such as indodicar-bocyanine �IDCC; �exc.: 650 nm; �em: 670 nm�, Cy5.5 ��exc:675 nm, �em: 694 nm� or indocyanine green �ICG, �exc:780 nm, �em: 830 nm� have been used to image specific mo-lecular markers of different diseases.16–19 In contrast to this,since one of the typical features of macrophages is theirphagocytotic activity, it should be possible to simply accumu-late fluorochrome in macrophages recruited within inflamma-tory areas, as this is likely to occur in peritonitis without theneed of highly sophisticated contrast agents. Whereas mostinvestigations deal with ex vivo labeling using magneticnanoparticles20,21 or fluorochromes,22 in vivo fluorochrometagging of cells has been only poorly considered for imaginguntil now.23–25 With the far-red and near-infrared optical meth-ods, a new noninvasive technique has entered the diagnosticfield in recent years.26–28 One particular advantage of thistechnique is that it can be operated with low costs, easy han-dling, and without radioactivity.

Therefore, using a novel hemicyanine dye DY-676 ��exc:676 nm; �em: 701 nm� with far-red/near-infrared fluoro-optical properties, we addressed the following issues:

• Whether it is possible to label macrophages and to detectthem with a whole-body small animal near-infrared imagingsystem in vitro by verifying macroscopic in vitro data usingflow cytometry and laser scanning microscopy

• Whether it is possible to label monocytes-macrophagesin vivo in a peritonitis animal model for diagnostic imagingwith the whole-body small animal near-infrared fluorescence�NIRF� imaging system.

2 Material and Methods2.1 Cell CultureCells from mouse macrophage cell line J774A.1 �Cell LinesService, Heidelberg, Germany� were cultured in DMEM �Dul-becco Modified Eagle medium; GIBCO BRL, Paisley, Scot-land� with 10% �v/v� FCS �fetal calf serum; GIBCO BRL,Paisley, Scotland� at 37°C in a 5% CO2 atmosphere and 95%humidity. To label 1.5�106 macrophages, cells were incu-bated for 3 h with the DY-676 �7.7 nmol/ml� or with culturemedium only as control for autofluorescence. After washing10 times with PBS �phosphate buffered saline, 100 mM, pH7.4� to remove noninternalized dye molecules, cells werescraped from the bottom of the culture dishes and resus-pended in PBS for further analysis.

2.2 In Vitro Far-Red/Near-InfraredFluorescence Imaging

Before harvesting, fluorochrome labeled cells were imagedwith a planar fluorescence reflectance system, the whole-bodysmall animal near-infrared fluorescence imager �bonSAI,Siemens, Erlangen, Germany�, using the 660/735-nm filtersfor excitation and emission at an acquisition time of 2.0 swith binning factor 2 �512�696 pixels matrix size�. Fluores-cence signals were detected with a charge-coupled devicecamera �pixel size 4.6�4.6 �m� and images �regions of in-terest, ROI� were semiquantitatively analyzed with theSYNGO software �Siemens, Erlangen, Germany�. Fluores-cence intensities �signal intensities in arbitrary units� were

presented as means±standard deviations. Comparisons ofmean fluorescence intensities between labeled and nonlabeledmacrophages were performed by means of Students t-test forindependent samples.

2.3 Flow Cytometry AnalysisIn vitro NIRF imaging of macrophages labeled for 24 h withDY-676 was further characterized with a FACSCalibur ana-lytical instrument using the fluorescence channel 3 �FL3,670-nm long pass �lp�� and the CellQuest software �BectonDickinson, Heidelberg, Germany�. For cell viability testing,cell samples were incubated with 2 �g/ml propidium iodide�Sigma-Aldrich, Steinheim, Germany� for 10 min and fluo-rescence emission was detected with the fluorescence channel2 �FL2, 585/42-nm bandpass �bp��.

2.4 Laser-Scanning MicroscopyIn order to further verify macrophage cell labeling, 105 cellsincubated with DY-676 �3 h, 7.7 �M� or with culture me-dium only were additionally sedimented and air-dried on His-tobond object slides �Marienfeld, Bad Mergentheim, Ger-many�, stained with 10 �g/ml 4�-6-diamidino-2-phenyl-indole �DAPI, Sigma-Aldrich, Steinheim, Germany�, and em-bedded with Permafluor �Immunotech, Marseille, France�.Confocal laser scanning microscopy was performed using aLSM510 system with the LSM510 image examiner software�Zeiss, Jena, Germany�.

2.5 In Vivo NIRF Imaging of Peritonitisin a Mouse Model

Nine 8-week-old male NMRI mice weighting approximately23 g were housed under standard conditions with food andwater ad libitum. Seven days prior to the start of experiments,all mice received a low pheophorbide diet �Altromin, Lage,Germany� in order to reduce autofluorescence of abdominalorgans. At the beginning of the experiments, all animals wereshaved on the ventral area. Animals were split into threegroups �i�, �ii�, and �iii�, with three animals per group. Perito-nitis was induced in mice of groups �i� and �ii� by intraperi-toneal injection of 3 mg of zymosanA �Sigma-Aldrich, Stein-heim, Germany� dissolved in 400 �l of PBS according toParente et al.29 Immediately after induction of peritonitis,mice from group �i� received DY-676 at a dose of 55 nmol offluorochrome per kg body weight into the tail vein. Group �ii�was not injected with fluorochrome as a control for autofluo-rescence of macrophages within the target region of diseasedmice. Mice from group �iii� got an intraperitoneal injection of400 �l of PBS �no peritonitis induction� followed by i.v. ap-plication of DY-676 as a control for unspecific accumulationof dye in the target region of nondiseased mice. All mice wereimaged directly before and at 6 h after �when monocyte-macrophage recruitment starts to increase, as reported byAjuebor et al.13� induction of peritonitis using the whole-bodysmall animal near-infrared fluorescence imaging system, asdescribed earlier. The acquisition time was 0.5 s with binningfactor 1 �1024�1392 pixels matrix size�. Statistical analysiswas performed as described earlier. All procedures were ap-proved by the regional animal committee and were in accor-dance with international guidelines on the ethical use ofanimals.

Lisy et al.: Diagnosis of peritonitis using near-infrared optical imaging…



2.6 Ex Vivo NIRF Imaging of Peritonitis6 h after peritonitis induction, the mice were sacrificed and exvivo whole-body NIRF imaging of the opened mice was per-formed. Afterward, the liver, spleen, and kidneys were with-drawn and imaged as well for assessment of whole-organfluorescence. The acquisition time was 0.5 s together with abinning factor of 1. Statistical analysis was performed as de-scribed earlier. Pieces of the small intestine were snap-frozen,and 6-�m thin sections were immunostained with DAPI and

Fig. 1 Biooptical NIRF recordings of in vitro labeled J774A.1 mac-rophages. Macroscopic light ��a� and �b�� and NIRF ��c� and �d�,660/730-nm exc./em� image of petri dishes covered with macro-phages after 3 h incubation with DY-676 �7.7 �M in DMEM� ��a� and�c�� or with DMEM medium only ��b� and �d�� as control for auto-fluorescence�. �e� Semiquantitative ROI-based evaluation of NIRF im-ages with four petri dishes per group: group 1 macrophages incubatedwith DY-676 �7.7 �M in DMEM�, and group 2 macrophages incu-bated with DMEM only. Data are expressed as means±standard de-viations. Differences between group 1 and control group 2 are signifi-cant �P�0.001�.

Fig. 2 Localization of DY-676 in in vitro labeled J774A.1 macro-phages. Upper row: Microscopic �CLSM� image of macrophages in-cubated for 3 h with DY-676 �7.7 �M in DMEM�: �a� DAPI for DNAstaining, �b� NIRF �650-nm lp�, and �c� merged image. Arrows showendosomal encapsulations of DY-676. Lower row: Macrophages incu-bated 3 h with DMEM medium only �control for autofluorescence�:�d� DAPI for DNA staining, �e� NIRF �650-nm lp�, and �f� mergedimage �scale bar=20 �m�.

Fig. 4 Whole body NIRF in vivo recordings of mice with inducedperitonitis. �a� to �f�: Macroscopic whole-body images of 8-week-oldNMRI mice �left: white light �0.3 s; binning factor 1�, right: NIRF,�0.5 s; binning factor 1��. �a� and �b� Mouse with peritonitis and i.v.application of DY-676 �55 nmol/kg body weight�. �c� and �d� Mousewith peritonitis, without DY-676. �e� and �f� Mouse without peritonitiswith i.v. application of DY-676 �55 nmol/kg�. �g� SemiquantitativeROI-based evaluation of NIRF images with three animals per group:�i� mice with peritonitis and i.v. application of DY-676 �55 nmol/kg�,�ii� mice with peritonitis without DY-676, and �iii� mice without peri-tonitis with i.v. application of DY-676 �55 nmol/kg�. Data are ex-pressed as means±standard deviations. Differences between group iand control groups ii and iii are significant �P�0.01�.




rat anti-mouse CD68 fluoresceinisothiocyanate �FITC�-conjugate �Biozol, Eching, Germany� or rat IgG2a FITC�Serotec, Oxford, UK� as isotype control, respectively, forconfocal laser scanning microscopy with the Zeiss LSM510system.

2.7 Isolation and NIRF Imaging of Peritoneal LavageIn order to verify that in vivo NIRF imaging was based on thepresence of fluorochrome labeled monocytes-macrophages atthe inflamed peritoneal region, all mice received a peritoneallavage immediately after sacrifice. Hereby, peritoneal cavitieswere washed with 5 ml of PBS containing 2 mM of ethylene-diaminetetraacetic acid �EDTA�. The cellular fraction was iso-lated by centrifugation. In order to assess the presence offluorochrome labeled cells, pellets containing 5�106 cellswere imaged with the whole-body small animal near-infraredimaging system as described earlier �acquisition time of 2.0 sand a binning factor of 2�. Afterward, cells from peritoneallavages were then fixated with 3% paraformaldehyde for15 min followed by careful washing and staining on ice withrat anti-mouse F4/80 FITC conjugate �Dianova, Hamburg,Germany� for 30 min. After extensive washing, cells wereanalyzed by flow cytometry using the FACSCalibur analyticalinstrument.

3 Results3.1 In Vitro NIRF Imaging of Labeled MacrophagesAnalyses with the whole-body small animal near-infraredfluorescence imaging system showed that fluorochrome la-beled macrophages grown on Petri dishes could be clearlyidentified. After incubation of cells for 3 h with DY-676�7.7 �M�, distinct fluorescence intensities were detected �Fig.1�a� and 1�c�� as compared to macrophages incubated withmedium only �Fig. 1�b� and 1�d��. Semiquantitative analysis

showed significantly increased �P�0.001� fluorescence in-tensities of labeled macrophages �1026±142 a.u.� as com-pared to controls �552±54 a.u., Fig. 1�e��.

The fact that macrophages were, indeed, labeled with thefluorochrome DY-676 was further verified by laser scanningmicroscopy. It was shown that DY-676 was stored inside thecells. Moreover, no homogeneous distribution inside the cellswas observed, but predominantly in compartments with highfluorescence intensity indicating an localized accumulation offluorochrome �Fig. 2�. Flow cytometry analysis revealed thatmacrophages treated with DY-676 could be clearly distin-guished from macrophages that have been incubated with me-dium only �Fig. 3�a� and 3�c��. As observed by fluorescence-activated cell sorting �FACS� analysis, more than 99% of allcells in the sample were labeled with the fluorochrome. FACSanalysis of propidium-iodide-stained cells revealed that bothsamples �with and without fluorochrome incubation� con-tained 85% viable �propidium iodide negative� and 15% non-viable cells �propidium iodide positive� after harvesting �Fig.3�b� and 3�d��.

3.2 In Vivo NIRF Imaging of Peritonitis

Whole-body small animal near-infrared fluorescence imagingshowed that 6 h after induction of peritonitis in mice thatreceived the fluorochrome DY-676 �55 nmol/kg, Fig. 4�a�and 4�b��, a distinct fluorescence area was depictable at theleft abdominal side, where zymosanA was previously in-jected. In comparison, control mice without peritonitis �injec-tion of PBS instead of zymosanA� that did receive DY-676 i.v.�Fig. 4�e� and 4�f�� showed a spot at the abdominal area witha comparatively much lower fluorescence intensity. Neverthe-less, mice having peritonitis that did not receive the contrastmedium DY-676 �control for autofluorescence of monocytes-macrophages within the target region of diseased mice, Fig.4�c� and 4�d�� revealed no fluorescence at all. Semiquantita-

Fig. 3 FACS analysis of in vitro labeled J774A.1 macrophages. Histogram blots of macrophages incubated 24 h with 7.7 nmol/ml DY-676 inDMEM ��a� and �b�� or DMEM only �control for autofluorescence� ��c� and �d��. Images �a� and �c� show relative fluorescence intensities in FL3channel �670-nm lp�; images �b� and �d� show relative fluorescence intensities in FL2 channel �585/42-nm bp� after 10-min propidium iodideincubation for staining of dead cells.




tive analysis of mice abdominal areas showed that 6 h afterinduction of peritonitis, fluorescence intensities of the ventralareas with peritonitis �group �i�, 869±87 a.u.� were increasedby a factor of 4 or higher compared to the data of both controlgroups �ii� �142±35 a.u.� and �iii� �188±24 a.u., Fig. 4�g��.

Fig. 5 Ex vivo NIRF imaging of macrophages from mice with andwithout peritonitis. �a� White-light image �0.3 s, binning factor 2� and�b� NIRF image �2.0 s; binning factor 2� of cell pellets from peritoneallavage from 8-week-old NMRI mice. Lower-right sample: cells frommouse with peritonitis with i.v. application of DY-676 �55 nmol/kg�.Upper-right sample: cells from mouse without peritonitis with i.v. ap-plication of DY-676 �55 nmol/kg, control for unspecific dye accumu-lation�. Left sample: cells from mouse with peritonitis without i.v. ap-plication of DY-676 �control for autofluorescence�. �c� Flowcytometric analysis of the cellular peritoneal infiltrate. The peritoneallavages of mice from group i were incubated with F4/80 FITC conju-gated mAb before FACS analysis. DotBlot shows relative fluorescenceintensities in FL3 channel �670-nm lp, DY-676� and relative fluores-cence intensities in FL1 channel �530/30-nm, bp, F4/80�.

Fig. 6 Whole-body NIRF ex vivo recordings of mice with inducedperitonitis. Macroscopic images �white light above �0.3 s; binning fac-tor 1� and NIRF below �0.5 s; binning factor 1�� of sacrificed8-week-old NMRI mice 6 h after i.v. injection of DY-676�55 nmol/kg�. �a� and �b� mouse with peritonitis, and �c� and �d�mouse without peritonitis �control for selective macrophage labeling�.

Fig. 7 Immunofluorescence staining of the small intestine of diseasedmice. �a� NIRF �red, �650-nm lp�, �b� CD68FITC, �c� merged withDAPI for DNA staining �blue�, and �d� rat IgG2a FITC isotype control�green� images derived from a mouse 6 h after injection of DY-676.Arrows show CD68+ monocytes-macrophages located in the laminapropria of villi within the mucosa containing DY-676. Images �e� NIRF�red, �650-nm lp�, �f� CD68FITC �green�, �g� merged with DAPI forDNA staining �blue�, and �h� rat IgG2a FITC isotype control �green�images derived from a diseased mouse without injection of DY-676�scale bars=50 �m�.




3.3 Ex Vivo NIRF Imaging of Peritoneal Lavage fromMice with and without Peritonitis

Analysis of cells isolated from peritoneal lavage using thewhole-body small animal near-infrared imaging system �Fig.5�a� and 5�b�� showing reaction cups containing cell sedi-ments� revealed the presence of fluorochrome labeled cellsonly in samples from mice with peritonitis having receivedthe contrast agent DY-676 �high fluorescence in the lower-right reaction cup, Fig. 5�b��. In contrast, isolated cells fromperitoneal lavage of mice with peritonitis but without contrastagent �left reaction cup, Fig. 5�b�� showed no fluorescenceintensity. The cell sediment in the upper-right reaction cupderived from mice that were only injected i.v. with DY-676but without induction of peritonitis revealed very weak signalintensities �Fig. 5�b��.

For the assessment of the constitution of the fraction offluorescent cells in the peritoneal lavage collected from miceof group �i�, the rat anti-mouse F4/80 FITC conjugate wasused as a marker for tissue macrophages. Flow cytometricanalysis revealed the presence of distinct populations on thebasis of forward and side scatter profiles �data not shown� aswell as on the basis of fluorescence emission �Fig. 5�c��. TheDotBlot clearly shows that cells with intermediate and highexpression levels of F4/80 antigen �28%� also revealed fluo-rescence intensities in the far-red/near-infrared range, whereasF4/80 negative cells �61%� did not. A third fraction of F4/80positive cells �11%� revealed no fluorescence in the far-red/near-infrared range.

3.4 Ex Vivo NIRF Imaging of Mice with andwithout Peritonitis

The whole-body NIRF images of opened mice with peritonitishaving received an i.v. injection of DY-676 showed areas ofdistinctly higher fluorescence intensities on the surface of theintestinal tract �Fig. 6�a� and 6�b�� as compared to mice with-out peritonitis but with i.v. injection of DY-676 �Fig. 6�c� and6�d��.

Immunostaining of the small intestine from diseased micethat were injected with DY-676 clearly showed that CD68-positive monocytes-macrophages located in the lamina pro-pria of villi within the mucosa revealed high fluorescence in-tensities in the far-red/near-infrared range �Fig. 7�a�–7�c��.Moreover, CD68-positive monocytes-macrophages betweenthe crypts were found to be intracellularly loaded with DY-676 �data not shown�. Whereas no fluorescence was observedusing the FITC-labeled isotype control, far-red/near-infraredfluorescent spots were still detected within the centers of thelamina propria �Fig. 7�d��. Immunostaining of the small intes-tine from diseased mice without injection of DY-676 showedno fluorescence intensities in the far-red/near-infrared range�Fig. 7�e�–7�h��.

Whole-organ fluorescence analysis �Fig. 8�a�–8�f�� showedthat some signal was detectable in the kidneys from mice bothwith and without peritonitis after application of the contrastmedium DY-676 �Fig. 8�d� and 8�f��. In comparison no fluo-rescence was observed in mice having received no DY-676�Fig. 8�d��. Semiquantitative analysis supported the aforemen-tioned findings �Fig. 8�g��. Particularly, low mean fluores-cence intensities of 210 to 270 a.u. were detected in the kid-neys of mice intravenously injected with DY-676, whereas the

mean fluorescence intensities of all other organs were evenlower than 100 a.u.

4 DiscussionOur results revealed that monocytes-macrophages could beefficiently labeled with DY-676 and imaged in vitro and invivo using a whole-body near-infrared fluorescence imagingsystem.

The in vitro experiments with mouse macrophage cell lineJ774A.1 showed that cells could be selectively imaged with awhole-body small animal imaging system after labeling withDY-676. Macrophage labeling with DY-676 was clearly veri-fied by FACS analysis; however, as compared to the afore-mentioned results, semiquantitative analysis of the images ob-tained from the whole-body small animal imaging systemshowed relatively higher fluorescence intensities of nonla-beled cells in relation to the labeled ones. These findings canbe attributed to the reflective effects of the culture dishes’smooth surface. Further on, FACS analysis using propidium

Fig. 8 Biooptical NIRF ex vivo recordings of organs from mice withinduced peritonitis. Macroscopic images �white light, above �0.3 s,binning factor 1� and NIRF below �0.5 s, binning factor 1�� of with-drawn organs from mouse with both peritonitis and i.v. application ofDY-676 �55 nmol/kg� ��a� and �d��; mouse with peritonitis without i.v.application of DY-676 ��b� and �e��; and mouse without peritonitiswith i.v. application of DY-676 �55 nmol/kg� ��c� and �f��. �g� Semi-quantitative ROI-based evaluation of NIRF images with three animalsper group: �i� mice with peritonitis and i.v. application of DY-676�55 nmol/kg�, �ii� mice with peritonitis without DY-676, and �iii� micewithout peritonitis with i.v. application of DY-676 �55 nmol/kg�. Dataare expressed as means±standard deviations.




iodide revealed that incubation with fluorochrome for 24 hdoes not affect cell viability. The histogram blots showed thatthere were only 15% of death cells both in DY-676 labeledand in nonlabeled samples. The presence of dead cells wasprobably caused by scraping off from the plastic matrix. Withrespect to the in vivo situation, this observation basically in-dicates that labeling of macrophages with fluorochromesshould not have an impact on cell viability. Moreover, DY-676 was predominantly stored within defined structures of thecytoplasm, which is presumably caused by the active uptakeinto organelles like endosomes during phagocytosis. The ex-act mechanisms for uptake of fluorochromes in macrophagesare not clear yet. Nevertheless, active phagocytosis of dyesafter binding to serum albumin has been already discussed24

in relation to the application of Cy5.5. Since the chemicalstructure of DY-676 is similar to that of Cy 5.5, it is conceiv-able that some unspecific binding of the dye to serum proteinstakes place. In this case, dye uptake by macrophages occursvia phagocytosis.

In vivo whole-body fluorescence imaging of mice at 6 hafter zymosanA-induced peritonitis showed that mice suffer-ing from this disease could be clearly distinguished from con-trols, indicating that, basically, an early specific identificationof the disease is possible using optical methods. Due to thefact that clearance of the fluorochrome in healthy mice takesbetween 4 and 6 h and as we intended to demonstrate feasi-bility to image early stages of inflammation, whenmonocytes-macrophages start to migrate to the target area, itwas necessary to inject the contrast agent immediately afterinduction of peritonitis. Other studies reported that monocyte-macrophage infiltration takes place for more than 24 h.11,13

Therefore, peritonitis could also be detected at later points intime, which has to be proven in further studies. Consideringthe clinical situation, a very effective monitoring of patients,e.g., during surgery, could be performed when monocytes-macrophages start to migrate to the target area, but furtherstudies are necessary to elucidate to what extent the charac-teristics of the peritonitis model used in the present study aretransferable to human beings.

All fluorescence data shown in the present study are basedon semiquantitative analysis of intrinsic values of the CCDcamera calculated by the SYNGO software. The influence ofinhomogeneous absorptive behaviour, reflection, or scatteringproperties30 of the samples or tissues investigated in this studycould not be considered. Nevertheless, as equivalent experi-mental conditions were used, the interpretation of semiquan-titative data is legitimate to show tendencies. The observeddifferences between the contrast in in vitro and in vivo experi-ments are not clear, but it is conceivable that the unspecificreflections in relation to the Petri dishes used in the presentstudy in the in vitro situation may have influenced the results.

The aforementioned in vivo observations are in agreementwith the results from ex vivo examinations showing that theapplication of the contrast medium led to a selective labelingof macrophages recruited at the target site. These results arereinforced, in principle, by the fact that the animal model usedis known to be characterized by a massive recruitment ofactivated monocytes-macrophages into the peritoneum at thearea of zymosanA injection, which increases between 6 h and24 h after induction of peritonitis.13 Presumably, in vivo up-

take of DY-676 by monocytes-macrophages occurs in wellvascularized mucosa layers before immigration into the in-flamed region takes place. This assumption is reinforced bythe observations of Ajuebor et al.,13 who reported a pro-nounced migration of monocytes-macrophages from the peri-toneum to the internal mucosa layers followed by a reap-pearence at the area of zymosanA injection. Further studiesshould consider and verify these relationships in more detailas well as identify to what extent they apply in connectionwith the situation in humans.

The presence of fluorochrome labeled cells in peritoneallavages extracted from the inflamed region could be con-firmed by analysis of NIRF images of cell sediments. In par-ticular, the strong fluorescence of the cell sediments derivedfrom the diseased mice after i.v. DY-676 application corre-sponds well to the in vivo results. The weak signal of theisolated cells from the mice without peritonitis that receivedonly contrast agent might be caused by the presence of la-beled monocytes-macrophages in consequence of the inter-vention itself, as demonstrated by the injection of harmlessPBS instead of zymosanA. As expected from the in vivo re-sults, no fluorescent cells could be found in cells from dis-eased mice that were not injected with the DY-676, indicatingthat autofluorescence emerging from monocytes-macrophagesper se or zymosanA itself had no effects on the observedfluorescence signals.

The fact that in vivo the fluorochrome DY-676 was mainlytaken up by monocytes-macrophages was demonstrated byFACS analysis. Hereby, cells expressing intermediate andhigh levels of F4/80 antigen were found to show fluorescenceintensities in the far-red/near-infrared range. Moreover, theseresults are consistent with previous reports13,31 showing thatthe F4/80 antigen is expressed on monocytes-macrophagesreappearing very early at the area of zymosanA injection up to24 h after onset of peritonitis. Further on, it was reported thatafter immunostaining of cells from peritoneal lavages againstthe F4/80 antigen, three subpopulations �lymphocytes—non-expressing; monocytes—intermediate levels; and macro-phages—high levels� were identified.13 In accordance withthat study, we state that monocytes-macrophages could beidentified as the fraction of cells in the peritoneal lavage col-lected containing the contrast agent DY-676. Moreover, thefact that in vivo whole-body fluorescence signals are based onfluorochrome labeled monocyte-macrophages is further rein-forced by immunofluorescence analysis of the small intestinefrom the area of zymosanA injection.

Whereas the results indicate that the detected fluorescencesignals are mainly associated with labeled cells, a nonspecificaccumulation of fluorochrome at the target region seems to beunlikely because controls injected with PBS only instead ofthe zymosanA suspension showed only very weak fluores-cence in vivo at the target site. This is supported by the ex vivoimages of opened control mice, where only very low fluores-cence intensities could be detected on the surface of the intes-tinal tract, due to the lack of an increased monocyte-macrophage population.

Compared to the fluorescence signals observed in the peri-toneal area of zymosanA injection, only very low signals weredetectable in the withdrawn highly vascularized organs, likethe kidneys, liver, and spleen. Particularly for the kidneys, thedetermined “out of body” fluorescence values are lower by a




factor of 3 in comparison to the results from the in vivo im-ages. Moreover, in vivo whole-body imaging of mice withoutperitonitis injected with DY-676 revealed no detectable fluo-rescence in the liver and kidneys. Therefore, the low signalintensities in these vascularized organs should not have con-tributed to the observed fluorescence intensity at the inflamedabdominal region.

The very low signals emitted from the kidneys suggest afast renal clearance mechanism of the fluorochrome thatwould be an advantage in relation to those macromolecularcontrast agents with a predominant hepatic deposition andmetabolism, like the clinically used indocyanine green32 orthe previously reported high-affinity contrast agents, wherefluorochrome molecules �e.g., Cy5.5� were coupled to macro-molecular antibodies �e.g., Herceptin� for tumor imaging.23,33

This is, to the best of our knowledge, the first time that peri-toneal macrophages have been efficiently labeled and imagedusing fluorochrome dyes.

Whereas the results of the present study indicate that peri-tonitis could be selectively detected by optical methods, oneof the possible limitations of this method consists of the lackof an adequate anatomical resolution that is associated withthe optical systems available up to now. The in vivo findingspresented here are based on planar fluorescence reflectanceimaging. No real tomographic information for exactly three-dimensional resolution of the inflammatory areas was avail-able. Nevertheless, different techniques like the very interest-ing approaches34 related to fluorescence mediated tomography�FMT� or the combination of NIRF imaging with other to-mographic devices should be valuable tools for providing fur-ther detailed information about the respective inflammatoryregion.

The present study shows that monocyte-macrophage cul-tures can be successfully labeled and imaged in their originalsize using a fluoro-optical contrast medium and that, accord-ingly, selective macrophage labeling and whole-body imagingcould also be performed in the in vivo situation, particularly inrelation to the detection of peritonitis as demonstrated in ananimal model. Furthermore, labeling of monocytes-macrophages with DY-676, as shown in the present study, aswell as using other potentially nontoxic NIRF fluorochromes,can be suitable for clinical diagnostics of other inflammatorydiseases. In that case, this technology would be limited todiseases that are located only a few centimeters below theskin, due to the limited penetration depth of far-red/near-infrared light. In our study, we intended to show the basicfeasibility using planar optical imaging whereby light penetra-tion into tissue is limited. However, based on the fact thatperitonitis is localized only a few cm under the skin surface, itis conceivable that peritonitis in humans could be detectedwith fluorescence-mediated tomography �penetration depth�10 cm; Ref. 34�. For more detailed in vivo imaging of de-veloping inflammatory processes, additional studies shouldfocus on specific target molecules such as cytokines and theirreceptors.

AcknowledgmentsThe authors would like to thank Dr. Harald Schubert �Institutefor Animal Research, University of Jena, Germany�, YvonneHeyne, and Brigitte Maron for excellent technical assistance.

References1. J. I. Greenfeld and C. M. Harmon, “Acute pancreatitis,” Curr. Opin.

Pediatr. 9, 260–264 �1997�.2. H. B. Reith, “Therapy of peritonitis today: surgical management and

adjuvant therapy strategies,” Langenbecks Arch. Chir. 382, S14–S17�1997�.

3. H. Ellis, “The hazards of surgical glove dusting powders,” Surg. Gy-necol. Obstet. 171, 521–527 �1990�.

4. J. Paavonen, “Pelvic inflammatory disease: from diagnosis to preven-tion,” Dermatol. Clin. 16, 747–756 xii �1998�.

5. R. W. Steiner and N. A. Halasz, “Abdominal catastrophes and otherunusual events in continuous ambulatory peritoneal dialysis patients,”Am. J. Kidney Dis. 15, 1–7 �1990�.

6. H. Wacha, T. Hau, R. Dittmer, and C. Ohmann, “Risk factors asso-ciated with intraabdominal infections: a prospective multicenterstudy: Peritonitis Study Group,” Langenbecks Arch. Chir. 384, 24–32�1999�.

7. V. Golash and P. D. Willson, “Early laparoscopy as a routine proce-dure in the management of acute abdominal pain: a review of 1,320patients,” Surg. Endosc 19, 882–885 �2005�, epub 12 May 2005.

8. A. Farooq and B. J. Ammori, “Laparoscopic diagnosis and manage-ment of primary bacterial peritonitis,” Surg. Laparosc. Endosc. Per-cutan. Tech. 15, 36–37 �2005�.

9. A. Korzets, Z. Korzets, G. Peer, J. Papo, D. Stern, J. Bernheim, andM. Blum, “Sclerosing peritonitis: possible early diagnosis by com-puterized tomography of the abdomen,” Am. J. Nephrol. 8, 143–146�1988�.

10. C. J. van der Laken, O. C. Boerman, W. J. Oyen, P. Laverman, M. T.van de Ven, F. H. Corstens, and J. W. van der Meer, “In vivo expres-sion of interleukin-1 receptors during various experimentally inducedinflammatory conditions,” J. Infect. Dis. 177, 1398–1401 �1998�.

11. S. J. Getting, R. J. Flower, L. Parente, R. de Medicis, A. Lussier, B.A. Woliztky, M. A. Martins, and M. Perretti, “Molecular determi-nants of monosodium urate crystal-induced murine peritonitis: a rolefor endogenous mast cells and a distinct requirement for endothelial-derived selectins,” J. Pharmacol. Exp. Ther. 283, 123–130 �1997�.

12. R. W. Kinne, R. Brauer, B. Stuhlmuller, E. Palombo-Kinne, and G.R. Burmester, “Macrophages in rheumatoid arthritis,” Arthritis Res.2, 189–202 �2000�.

13. M. N. Ajuebor, R. J. Flower, R. Hannon, M. Christie, K. Bowers, A.Verity, and M. Perretti, “Endogenous monocyte chemoattractantprotein-1 recruits monocytes in the zymosan peritonitis model,” J.Leukoc Biol. 63, 108–116 �1998�.

14. K. Konig, “Multiphoton microscopy in life sciences,” J. Microsc.200, 83–104 �2000�.

15. Z. Gryczynski, I. Gryczynski, and J. R. Lakowicz, “Fluorescence-sensing methods,” Methods Enzymol. 360, 44–75 �2003�.

16. A. Becker, C. Hessenius, K. Licha, B. Ebert, U. Sukowski, W. Sem-mler, B. Wiedenmann, and C. Grotzinger, “Receptor-targeted opticalimaging of tumors with near-infrared fluorescent ligands,” Nat. Bio-technol. 19, 327–331 �2001�.

17. K. Licha, C. Hessenius, A. Becker, P. Henklein, M. Bauer, S. Wis-niewski, B. Wiedenmann, and W. Semmler, “Synthesis, characteriza-tion, and biological properties of cyanine-labeled somatostatin ana-logues as receptor-targeted fluorescent probes,” Bioconjugate Chem.12, 44–50 �2001�.

18. R. Weissleder and U. Mahmood, “Molecular imaging,” Radiology219, 316–333 �2001�.

19. K. Licha, B. Riefke, V. Ntziachristos, A. Becker, B. Chance, and W.Semmler, “Hydrophilic cyanine dyes as contrast agents for near-infrared tumor imaging: synthesis, photophysical properties and spec-troscopic in vivo characterization,” Photochem. Photobiol. 72, 392–398 �2000�.

20. M. Lewin, N. Carlesso, C. H. Tung, X. W. Tang, D. Cory, D. T.Scadden, and R. Weissleder, “Tat peptide-derivatized magnetic nano-particles allow in vivo tracking and recovery of progenitor cells,” Nat.Biotechnol. 18, 410–414 �2000�.

21. U. Himmelreich, R. Weber, P. Ramos-Cabrer, S. Wegener, K. Kandal,E. M. Shapiro, A. P. Koretsky, and M. Hoehn, “Improved stem cellMR detectability in animal models by modification of the inhalationgas,” Mol. Imaging 4, 104–109 �2005�.

22. R. W. Dirks, C. Molenaar, and H. J. Tanke, “Visualizing RNA mol-ecules inside the nucleus of living cells,” Methods 29, 51–57 �2003�.




23. B. Ballou, G. W. Fisher, A. S. Waggoner, D. L. Farkas, J. M. Reiland,R. Jaffe, R. B. Mujumdar, S. R. Mujumdar, and T. R. Hakala, “Tumorlabeling in vivo using cyanine-conjugated monoclonal antibodies,”Cancer Immunol. Immunother 41, 257–263 �1995�.

24. A. Hansch, O. Frey, I. Hilger, D. Sauner, M. Haas, D. Schmidt, C.Kurrat, M. Gajda, A. Malich, R. Brauer, and W. A. Kaiser, “Diagno-sis of arthritis using near-infrared fluorochrome Cy5.5,” Invest. Ra-diol. 39, 626–632 �2004�.

25. X. Dong, S. Swaminathan, L. A. Bachman, A. J. Croatt, K. A. Nath,and M. D. Griffin, “Antigen presentation by dendritic cells in renallymph nodes is linked to systemic and local injury to the kidney,”Kidney Int. 68, 1096–1108 �2005�.

26. R. R. Alfano, S. G. Demos, and S. K. Gayen, “Advances in opticalimaging of biomedical media,” Ann. N.Y. Acad. Sci. 820, 248–271�1997�.

27. S. Andersson-Engels, C. Klinteberg, K. Svanberg, and S. Svanberg,“In vivo fluorescence imaging for tissue diagnostics,” Phys. Med.Biol. 42, 815–824 �1997�.

28. J. C. Hebden and D. T. Delpy, “Diagnostic imaging with light,” Br. J.Radiol. 70, S206–S214 �1997�.

29. M. Perretti, E. Solito, and L. Parente, “Evidence that endogenousinterleukin-1 is involved in leukocyte migration in acute experimentalinflammation in rats and mice,” Agents Actions 35, 71–78 �1992�.

30. V. Ntziachristos, J. Ripoll, L. V. Wang, and R. Weissleder, “Lookingand listening to light: the evolution of whole-body photonic imag-ing,” Nat. Biotechnol. 23, 313–320 �2005�.

31. J. F. Cailhier, M. Partolina, S. Vuthoori, S. Wu, K. Ko, S. Watson, J.Savill, J. Hughes, and R. A. Lang, “Conditional macrophage ablationdemonstrates that resident macrophages initiate acute peritoneal in-flammation,” J. Immunol. 174, 2336–2342 �2005�.

32. S. G. Ketterer, B. D. Wiegand, and E. Rapaport, “Hepatic uptake andbiliary excretion of indocyanine green and its use in estimation ofhepatic blood flow in dogs,” Am. J. Physiol. 199, 481–484 �1960�.

33. I. Hilger, Y. Leistner, A. Berndt, C. Fritsche, K. M. Haas, H. Kos-mehl, and W. A. Kaiser, “Near-infrared fluorescence imaging ofHER-2 protein over-expression in tumour cells,” Eur. Radiol. 14,1124–1129 �2004�.

34. V. Ntziachristos, C. Bremer, and R. Weissleder, “Fluorescence imag-ing with near-infrared light: new technological advances that enablein vivo molecular imaging,” Eur. Radiol. 13, 195–208 �2003�.




diagnosis of peritonitis using near-infrared optical ... filediagnosis of peritonitis using...

Documents